Palo Alto Networks agrees to US$4.5 billion Applied Compute acquisition
The all-stock transaction would expand Palo Alto’s AI business, with regulatory clearance still required and closing expected in fiscal 2027.

Palo Alto Networks has agreed to acquire Applied Compute for US$4.5 billion in stock, expanding into technology that helps businesses train and operate customized open-weight AI models. The transaction requires regulatory approval, with completion expected during Palo Alto’s fiscal third quarter of 2027.
Applied Compute was established by three former OpenAI researchers. Its co-founder and chief executive, Yash Patil, is set to join Palo Alto as co-chief AI officer once the acquisition closes. The startup said its run-rate revenue had reached US$100 million.
Palo Alto plans to offer businesses tools to train, modify and operate open-weight models using public cloud infrastructure or their own sovereign environments. Such models allow users to download and alter the numerical weights that govern their behavior. Applied Compute will continue serving existing customers, while Palo Alto plans to market the platform to its customer base.
The company expects the acquisition to contribute more than US$1 billion in annual recurring revenue by fiscal 2029. It also plans more than 100 megawatts of inference capacity, with associated capital spending below US$3 billion through fiscal 2030.
Palo Alto announced the acquisition alongside fiscal first-quarter earnings. For the three months ending 31 October, revenue increased 39% to US$3.44 billion, while net income rose to US$482 million from US$334 million a year earlier.
The company also increased its full-year forecasts, projecting revenue between US$14.52 billion and US$14.57 billion and adjusted per-share earnings between US$4.26 and US$4.30.




